| name | reporting |
| description | Generate publication-quality figures and markdown reports for biomedical ML experiments. Use when: (1) Creating heatmaps, forest plots, radar charts, or sensitivity-specificity scatter plots, (2) Plotting ROC curves, precision-recall curves, or Kaplan-Meier survival curves, (3) Writing structured experiment reports, (4) Exporting results to LaTeX tables for paper submission, (5) Producing figures meeting publication DPI and formatting standards.
|
Scientific Reporting
Workflow
Generating a report involves these steps:
- Choose figure types -- based on what dimensions you're comparing
- Generate figures -- at appropriate DPI and formatting
- Export tables -- to markdown and/or LaTeX
- Write report -- structured markdown with figures and tables
- Save to dated directory --
docs/<YYYYMMDD>/ to prevent overwrites
Decision Tree
What are you visualizing?
What is the target venue?
- Internal review → DPI 150, PNG format
- Journal/conference submission → DPI 300+, PDF/SVG format
- ASK the user about target venue and formatting requirements
ASK the user before starting:
- What figures do they need?
- Target venue (determines DPI, formatting)?
- Color-blind-friendly palette needed?
- What file format (PNG, PDF, SVG)?
References
| File | Read When |
|---|
| references/reporting-patterns.md | Heatmaps, forest plots, radar/spider charts, sensitivity-specificity scatter, dated output directories, markdown report template, figure quality guidelines |
| references/curve-plots.md | ROC curves (single and with bootstrap CI), precision-recall curves, Kaplan-Meier survival curves with log-rank test |
| references/latex-tables.md | Exporting results to LaTeX tables with booktabs formatting, standalone preview files |
